Price information collected online and short-term inflation forecasts 1
|
|
- Christal Jacobs
- 6 years ago
- Views:
Transcription
1 IFC-Bank Indonesia Satellite Seminar on Big Data at the ISI Regional Statistics Conference 2017 Bali, Indonesia, 21 March 2017 Price information collected online and short-term inflation forecasts 1 Isaiah Hull, Marten Löf and Markus Tibblin, Sveriges Riksbank 1 This paper was prepared for the meeting. The views expressed are those of the authors and do not necessarily reflect the views of the BIS, the IFC or the central banks and other institutions represented at the meeting.
2 Price information collected online and short-term inflation forecasts Isaiah Hull, Mårten Löf and Markus Tibblin 1 Abstract Forecasting short-term inflation developments (e.g. inflation over the coming months) is important for a central bank. There are certain elements within the published inflation figures that are volatile and inherently hard to forecast even in the short-run. Fruit and vegetable prices, energy prices and air travel prices are examples of product groups within the inflation measure that historically have held a high degree of volatility in Sweden. An automatic internet data collection process was developed to collect sales prices daily for selected fruits and vegetables from a number of Swedish online retailers. The results indicate that the information from the daily data could increase the precision in short-term inflation forecasts in Sweden. Keywords: inflation, forecasting, big data JEL classification: L11, L81, E30, C55 1. Short-term inflation forecasts matter to central banks The objective for monetary policy at the Riksbank, like many other central banks, is to maintain price stability. This is interpreted as keeping inflation, i.e. consumer prices measured through an index based on a basket of goods and services households tends to consume, low and stable. 2 The consumer price index (CPI) consists of a range of sub-categories. In Sweden, fruit and vegetables, air travel and fuel prices are examples of sub-categories, which historically have shown a relatively high level of price volatility. This volatility may at times create challenges in forecasting CPI in the short-term (e.g. the coming months) although these sub-categories sum to only a small share of total CPI. Short-term CPI volatility and related short-term forecast errors may seem like a minor issue, given the fact that monetary policy decisions are based on inflation development over a longer time horizon. However, accurate forecasts of future inflation are dependent on precise information regarding current inflation, as well as good forecasts of short-term inflation developments. It is important from a forecasting, and also a monetary policy, perspective to be able to decide to what extent a large deviation between forecast and outcome is due to temporary or more permanent factors. A large forecast error that is due to permanent 1 The authors work at the Monetary Policy Department of the Riksbank. The opinions here are the sole responsibility of the authors and should not be viewed as reflecting the views of the Riksbank. The authors would like to thank colleagues at the Riksbank for their valuable comments on previous drafts. 2 Price information collected online and short-term inflation forecasts 1
3 factors can affect the path of the new inflation forecast over the next twelve months, assuming that no other changes will occur, see Figure 1. Inflation outcome and forecast revision due to short-term forecast error Yearly percentage change Figure 1 Note: The figure illustrates how a forecast error at time t (difference between the solid and the dashed red line) may affect the forecast path for inflation in the coming year (blue dashed line). 2. New data and new analytical methods The information published on the Internet is growing very fast. It has been widely discussed how, for example, central banks can make use of new methods to collect and analyse this type of data. The Riksbank arranged a Big Data workshop in September 2015, at which a number of central banks, researchers and private firms laid out examples where new data and methods arising from Big Data may support analysis and decision-making at central banks. 3 One area of growing interest among central banks is the increasing amount of sales price information available online. Leading research in this area has been conducted in the Billion prices project. This project was initiated by Cavallo and Rigobon, who have in a number of studies shown that a price index produced using online price data follows the official measures of consumer prices such as CPI fairly well (Cavallo and Rigobon, 2014). Also, through discussions with retailers in relation to the Riksbank s Business survey it is clear that there are in general small differences in consumer prices between offline and online prices in Sweden, 4 which is also in line with a large international comparison study conducted by Cavallo (2016). Furthermore, it has been shown that online pricing data /rap_foretagsundersokning_160615_eng.pdf 2 Price information collected online and short-term inflation forecasts
4 performs well in forecasting CPI produced by national statistical offices and for some countries outperform models that include offline data (Aparicio and Bertolotto, 2016). 3. The pilot project collecting online fruit and vegetable prices from the internet Utilising online price data may potentially overcome a number of current issues in short-term inflation forecasts. First, there are no time lags in data collection, as online prices can be collected in real-time i.e. price data are available before official inflation figures are published. Moreover, collecting price data online generates information at a very granular level. This enable a more detailed analysis, which for example could give information on whether an unusual price development is temporary or not. Also, as online data collection can be automated, daily price indices can be produced and included in models at a low cost and with very limited resources. A small pilot study was initiated to investigate whether prices of fruit and vegetables that are available online could improve the accuracy of short-term inflation forecasts in Sweden. Prices of fruit and vegetables make up 3 percent of the CPI basket. Although these prices constitute a small part of the basket, strong price movements in this sub-index can have a clear impact on the aggregated figures. The pilot project was set up with the aim to: 1. Create a process for automatic online data collection collecting online price data daily for a few selected fruit and vegetables from Swedish retailers with e-commerce. Online prices have been collected for oranges, bananas, peppers, apples, cucumbers, cabbage, grapes, cauliflowers, pears, leeks and tomatoes. The correlation between the subset of selected fruit and vegetable prices and the total index of fruit and vegetables in the CPI is fairly high, see Figure 2 below. Hence, the subset of prices captures the variation in the total index for fruits and vegetables in the CPI quite well. It was therefore assumed that internet data for selected products would be useful when forecasting the overall CPI-index for fruit and vegetables. 2. Create a weighted monthly price index based on observed price changes online for the selected fruit and vegetables mentioned in Test whether the constructed online price index adds value to the existing shortterm forecasting models for fruit and vegetables currently used by the Riksbank. The rest of the paper is outlined as follows: First a brief description of the data collection method is given, then the index construction and data transformations are described. Finally, the results are presented, including a discussion and overall conclusions drawn so far from the pilot project. Price information collected online and short-term inflation forecasts 3
5 Prices of selected products in CPI and the aggregate of fruit and vegetables in the CPI Monthly percentage change Figure 2 4. Scraping We use a technique called scraping to collect the data for this project. Scraping involves sending programmatic requests to a website s server. The server responds by returning the underlying code that would be executed in a browser if you were to visit a given page on its website. The code s structure can then be employed to identify targeted items, such as product names and prices. We use additional tools to schedule the scraping tasks so that they are automatically performed on the targeted sites at the same time each day. We also limit the speed at which we send requests to each website to ensure that no strain is placed on their servers. All data collection tasks are performed on a Linux virtual private server (VPS). The server executes three scripts in sequence at the same time each day, as illustrated by Figure 3. The first visits the websites of four large grocery retailers. It extracts the code from all pages related to fruit and vegetables. A list of all product prices and names is then identified in the code and saved in a.csv file for each location and day. The raw code is also saved in.txt format for 90 days, allowing us to correct errors discovered later. The script then uses regular expressions, which identify patterns in text to filter the data, creating a second.csv file for each location and day that consists only of targeted fruits and vegetables. After the scrape is finished, the server executes the second script, which merges the filtered location and day files with past data. It first identifies matches between products in the new data and products identified in previous scrapes. Items that were 4 Price information collected online and short-term inflation forecasts
6 not identified in previous scrapes are assigned new location-product IDs and products that already have IDs are merged with existing location-product time series. Finally, the server executes the third script, which checks for errors. This script identifies the number of files produced, the size of those files, and the data types used in those files. The server then delivers an error report via SMS, allowing us to quickly identify and correct errors. The daily scraping process Figure 3 5. Forming indices In a first step the collected online prices are sorted into different groups, orange prices in one group, and apple prices in another group and so on. The dataset is then truncated so that it exactly matches the measurement weeks in the Swedish CPI survey. 5 In the next step a geometric mean is calculated for each product and month, i.e. an average price of oranges in January, one for orange prices in February and so on. 5 Price data on fruit and vegetables in the CPI were collected for three weeks in the middle of each month (a total of 21 days) until December Thereafter, Statistics Sweden has made some changes. Price information collected online and short-term inflation forecasts 5
7 Figure 4 shows the monthly percentage changes in prices for oranges, cucumbers, peppers, and tomatoes from May 2015 to March The blue lines show monthly percentage price changes according to the CPI, while the red lines show the corresponding changes based on information from the internet. The correlation between the CPI prices and prices available online are relatively high. The correlation is highest for prices on cucumbers and tomatoes. The lowest correlations are measured for prices on oranges. Price changes according to internet data and corresponding price changes according to the CPI Monthly percentage change May 2015 March 2017 Figure 4 Note: The red lines indicate online data while the blue lines indicate data from official CPI These price changes are then merged in to an overall index (henceforth denoted the online pilot price index) using the CPI weights according to CPI for the fruit and vegetable prices collected online. The left panel of Figure 5 compares the online pilot price index (red line) with a weighted CPI-index for the corresponding fruit and vegetable prices (blue line). The right panel in Figure 5 shows the online pilot price index together with the total price index for fruit and vegetables in the CPI (i.e. not only the fruit and vegetables collected in the pilot study). Naturally, the correlation decreases when comparing the online pilot price index with the total index for fruit and vegetables (left panel compared to right panel in figure 5). However, it is still fairly high and the online pilot price index captures most of the volatility in the official fruit and vegetables index. 6 Price information collected online and short-term inflation forecasts
8 Comparison between an aggregated index based on internet data and two different indexes with CPI data Monthly percentage change May 2015 March 2017 Figure 5 Note: The figure show aggregated price changes according to the data from the internet (red lines). Blue lines show price changes according to CPI. The blue line in the left hand panel show price changes if one use the same products as in the internet collection. The blue line in the right panel shows price changes according to the index for fruit and vegetables in CPI. 6. Forecast evaluation In this section we want to test more formally whether the internet prices can be used when forecasting the index for fruit and vegetables in the CPI. Here we compare the forecasts from the Riksbank's current models for fruit and vegetables with the forecasts generated using the internet prices. A variety of indicators are included in the models now used by the Riksbank for short-term forecasting. The models for prices of fruit and vegetables include one indicator at a time, together with a moving average of an exchange rate index (KIX). In addition, lags of the dependent variable, dummies for outliers and in some cases moving average terms are included in the models. The results are summarized as the mean forecast from all of these specifications. This approach is denoted MEAN in the evaluation below. Principal component analysis is also used to summarise the information from the indicators in the first step. The resulting Principal component indices can be seen as a weighted averages of all the indicators. In that case the models include one or more of these summary indices instead of the individual indicators. This approach is denoted PC in the evaluation below. These model-based forecasts are compared with an approach using the online price index. Here we simply use the online price index as a forecast for the aggregate of fruit and vegetables in the CPI. We denote this second approach (OP). The evaluation period is May 2015 to March The root mean square error (RMSE) is used to compare the forecasting ability between the approaches. The RMSE summarises the standard deviation in the forecast errors and their systematic deviation. The lower the estimated RMSE, the better the forecasting ability. A forecast that is always correct has a zero RMSE. Table 1 below summarizes the results. During this short evaluation period the best approach has been OP, where the online price index are used directly. Hence, it seems Price information collected online and short-term inflation forecasts 7
9 like online price information could add value in forecasting price changes for fruit and vegetables in the coming month. However, the evaluation period is very short, and longer time series are required to draw clearer conclusions. RMSE for different approaches, May 2015-March Table 1 Nowcasting procedures Online prices Horizon Mean PC OP 1 month Discussion The work on developing scrapers for automatically collecting and analysing online price data has been a bit of a trial-and-error exercise. The codes developed have had to be tailored to handle a range of different, changing website layouts. However, it has been possible to build scrapers coping with changing layouts, thus making it possible to maintain a collection process with very limited resources. The journey of analysing online micro data has just started during this pilot project. A fair amount of time has been devoted to investigating whether further transformation of the online data could increase forecasting ability. For example, an index only based on online prices showing high correlation with corresponding official CPI prices was constructed. The online price information has also been summarized using Principal Component Analysis. These types of indices have at times performed very well, but also shown large variability and overall not been as good as just the simple online index described above. Thus, no transformation is currently performed on the data. This could potentially be an area of further investigation. Additionally, the collected daily data could also be analysed from the perspective of firms pricing behaviour. Detailed micro data may be a source of understanding when, why and how firms change their prices. This could also be a potential area of further investigation. 8. Conclusions The pilot project has revealed a number of insights regarding online price collection and the analysis of such data in relation to inflation forecasting. First of all, it has been proved possible to consistently scrape online retail prices from e-commerce websites held by retailers in Sweden. On the one hand, creating scripts and IT-processes ensuring a stable data collection takes time and requires programming competence not traditionally available at central banks. On the other hand, once scrapers have been put in place, only minimum maintenance and development have been required. Furthermore, the result so far indicates that online pricing data add some value when forecasting short-term developments of consumer prices for fruit and vegetables. 8 Price information collected online and short-term inflation forecasts
10 Given this outcome, there may be scope for further expanding the collection of prices available online and using them as input in short-term forecasting models. The aggregated time series of online prices is still short and further analysis is required to ensure that the online price index have forecasting ability also in the future. The result also indicates that there may be room for increasing the collection of prices available on the internet and using them in short-term forecast models. However, there are still many questions that should be investigated, such as how to use the information in the best possible way. Price information collected online and short-term inflation forecasts 9
11 References Aparicio, Diego, & Manuel Bertolotto. (2016). Forecasting Inflation with Online Prices. Working Paper - MIT. Bernanke, B. S. & Boivin, J. (2003), Monetary policy in a data-rich environment, Journal of Monetary Economics 50 (3), Bertoloto, M. & Cavallo, A. & Rigobon, R. (2014), "Using Online Prices to Anticipate Official CPI Inflation," UTokyo Price Project Working Paper Series 031, University of Tokyo, Graduate School of Economics. Cavallo, F. Alberto, (2016), "Are Online and Offline Prices Similar? Evidence from Large Multi-Channel Retailers," NBER Working Papers 22142, National Bureau of Economic Research, Inc. Stock, J. H. & Watson, M.W. (2002), Forecasting using principal components from a large number of predictors, Journal of the American Statistical Association 97:460, Stock, J. H. & M. W. Watson, (2004), Combination forecasts of output growth in a seven- country data set, Journal of Forecasting 23 (Issue 6), Price information collected online and short-term inflation forecasts
12 IFC-Bank Indonesia Satellite Seminar on Big Data at the ISI Regional Statistics Conference 2017 Bali, Indonesia, 21 March 2017 Scraped sales price information and short-term CPI forecasts 1 Isaiah Hull, Marten Löf and Markus Tibblin, Sveriges Riksbank 1 This presentation was prepared for the meeting. The views expressed are those of the authors and do not necessarily reflect the views of the BIS, the IFC or the central banks and other institutions represented at the meeting.
13 Scraped Sales Price Information and Short-Term CPI Forecasts Isaiah Hull, Mårten Löf, and Markus Tibblin Sveriges Riksbank March 3, 2017 Scraped Sales Price Information and Short-Term CPI Forecasts 0/11
14 Background Does Sales Price Information Scraped from the Internet Increase the Precision of Short-Term Inflation Forecasts? Scraped Sales Price Information and Short-Term CPI Forecasts 1/11
15 Background Pilot project started in December 2014 Collected price data from Swedish grocery retailers with online presence Brick-and-mortar stores Internet-only retailers Multiple store-locations for largest retailer Constructed indices that are used as input for short-term inflation forecast Scraped Sales Price Information and Short-Term CPI Forecasts 2/11
16 Background Focused on fruits and vegetables High price variation over time Difficult to forecast Limited scope to subset with known CPI weights oranges, apples, bananas, cucumbers, peppers, tomatoes, pears, cabbage Scraped Sales Price Information and Short-Term CPI Forecasts 3/11
17 Methods A server runs three scripts in sequence daily The first script visits a number of grocery retailers It identifies and collects code associated with all fruits and vegetables on the website Scraped Sales Price Information and Short-Term CPI Forecasts 4/11
18 Methods The script parses the code to extract product names and prices It also applies a filter to generate a second file that contains only targeted fruits and vegetables The second script merges all daily scrape files and updates time series Scraped Sales Price Information and Short-Term CPI Forecasts 5/11
19 Methods Finally the server executes third script that scans all new files to determine whether there were any errors The first script also maintains a 90-day rolling archive of the raw code extracted from the website, so that revisions can be made if any errors are discovered at a later date Scraped Sales Price Information and Short-Term CPI Forecasts 6/11
20 Results Filtered dataset Data stored using Dropbox 3000 price series Series means first computed for selected time window Means computed across items of same type Scraped Sales Price Information and Short-Term CPI Forecasts 7/11
21 Results Calculate averages for price series using CPI weights: index = w cpi oranged(p orange ) w cpi tomato d(p tomato) (1) Use index as input to short term CPI forecast Scraped Sales Price Information and Short-Term CPI Forecasts 8/11
22 Results CPI vs. Scraped Data Scraped Sales Price Information and Short-Term CPI Forecasts 9/11
23 Results CPI vs. (Adjusted) Scraped Data Scraped Sales Price Information and Short-Term CPI Forecasts 10/11
24 Summary Online price collection adds value to current CPI forecasts Forecast error reduced Must perform additional cross validation Constructed robust, low-maintenance, scalable system for price collection No major code changes required since completion Maintenance infrequent Scraped Sales Price Information and Short-Term CPI Forecasts 11/11
THE ROLE OF NESTED DATA ON GDP FORECASTING ACCURACY USING FACTOR MODELS *
Andrejs Bessonovs University of Latvia, Latvia THE ROLE OF NESTED DATA ON GDP FORECASTING ACCURACY USING FACTOR MODELS * Abstract The paper studies an impact of nested macroeconomic data on Latvian GDP
More informationPublicly-Shared Datasets from The Billion Prices Project
Publicly-Shared Datasets from The Billion Prices Project Alberto Cavallo MIT & NBER IFM Data Project NBER Summer Institute 2015 BPP Online Data Online prices collected Directly from each retailer s website
More informationOnline Appendix for Are Online and Offline Prices Similar? Evidence from Multi-Channel Retailers
Online Appendix for Are Online and Offline Prices Similar? Evidence from Multi-Channel Retailers Alberto Cavallo MIT & NBER This version: August 29, 2016 A Appendix A.1 Price-level Comparison with Amazon.com
More informationThe Billion Prices Project Research and Inflation Measurement Applications
The Billion Prices Project Research and Inflation Measurement Applications Alberto Cavallo MIT June 2014 Micro Price Data in Macroeconomics Data Sources Statistical Offices (CPI, PPI, IPI) Scanner Data
More informationn Economic Commentaries
n Economic Commentaries Our view of resource utilisation plays a central role in the analysis of how monetary policy should be formulated. However, measuring resource utilisation is difficult in several
More informationBarbro Wickman-Parak: The Riksbank and monetary policy perspectives on interest rate forecasts
Barbro Wickman-Parak: The Riksbank and monetary policy perspectives on interest rate forecasts Speech by Ms Barbro Wickman-Parak, Deputy Governor of the Sveriges Riksbank, at a Statistics Sweden seminar,
More informationThe Billion Prices Project Big Data in Macroeconomics
The Billion Prices Project Big Data in Macroeconomics Alberto Cavallo MIT & NBER Symposium Building the Financial System Harvard Law - December 2017 Based on Cavallo & Rigobon (2016) The Billion Prices
More informationThe Billion Prices Project: Building Economic Indicators From Online Data. Alberto Cavallo MIT Sloan
The Billion Prices Project: Building Economic Indicators From Online Data Alberto Cavallo MIT Sloan UNECE CPI Meeting, Geneva, May 31 st 2012 STATE STREET ASSOCIATES Our approach: use online data to build
More informationTHE RESPONSES OF INTERNET RETAIL PRICES
THE RESPONSES OF INTERNET RETAIL PRICES TO AGGREGATE SHOCKS: A HIGH-FREQUENCY APPROACH Yuriy Gorodnichenko UC Berkeley & NBER Viacheslav Sheremirov FRB Boston Oleksandr Talavera Swansea U November 2017
More informationA NEW COINCIDENT INDICATOR FOR THE PORTUGUESE PRIVATE CONSUMPTION*
A NEW COINCIDENT INDICATOR FOR THE PORTUGUESE PRIVATE CONSUMPTION* António Rua** 1. INTRODUCTION As it is well known, private consumption plays a major role in overall economic activity (1). Therefore,
More informationNOWCASTING PRIVATE CONSUMPTION: TRADITIONAL INDICATORS, UNCERTAINTY MEASURES, AND THE ROLE OF INTERNET SEARCH QUERY DATA
NOWCASTING PRIVATE CONSUMPTION: TRADITIONAL INDICATORS, UNCERTAINTY MEASURES, AND THE ROLE OF INTERNET SEARCH QUERY DATA María Gil, Javier J. Pérez, and Alberto Urtasun Banco de España A. Jesus Sánchez
More informationCONSTRUCTING AN ECONOMIC COMPOSITE INDICATOR FOR THE UAE 1
CONSTRUCTING AN ECONOMIC COMPOSITE INDICATOR FOR THE UAE 1 ASSIL EL MAHMAH Research and Statistics Department June 2017 1 The views expressed in this paper are those of the author and should not be interpreted
More informationNICE Customer Engagement Analytics - Architecture Whitepaper
NICE Customer Engagement Analytics - Architecture Whitepaper Table of Contents Introduction...3 Data Principles...4 Customer Identities and Event Timelines...................... 4 Data Discovery...5 Data
More informationScraped Dat a and St icky Prices
Scraped Dat a and St icky Prices Albert o Cavallo MIT & NBER Banque de France, December 2015 Motivation St icky prices are a fundament al element of t he monet ary transmission mechanism in many macro
More informationSession IV Improving the quality of price statistics: Monitoring consumer price trend using daily price data of online grocery stores in India
Session IV Improving the quality of price statistics: Monitoring consumer price trend using daily price data of online grocery stores in India Anirban Sanyal Reserve Bank of India Content I. Research question
More informationStatus update on Scanner and online data integration projects
Status update on Scanner and online data integration projects Ken Van Loon Statistics Belgium UN GWG on Big Data for Official Statistics Training workshop on scanner and on line data 6 7 November 2017
More informationBusiness Intelligence. Get a clear picture of your business performance
Business Intelligence Get a clear picture of your business performance Copyright 2018 Pronto Software Limited. All rights reserved. 740 Solutions Overview: Business Intelligence Trademarks PRONTO, PRONTO
More informationAn introduction to scanner data methods
An introduction to scanner data methods UN GWG on Big Data for Official Statistics Training workshop on scanner and on-line data 6-7 November 2017 Bogota, Colombia Presented by: Michael Holt (michael.holt@abs.gov.au)
More informationCase Study: Netlife. Timeline. 85 employees 2016 Started to use CV Partner > 50% less time spent on CVs
Case Study: Netlife Netlife started using CV Partner in 2016. Netlife use CV Partner for both keeping track of internal resources and competence, as well as creating tailored bids for public and private
More informationDemystifying big data in official statistics it is not rocket science!
Demystifying big data in official statistics it is not rocket science! Jens Mehrhoff, Eurostat Second Statistics Conference Banco Central de Chile Santiago, 3 4 October 2017 Eurostat Structure of the presentation
More informationEnergy Advisor. Introduction. Energy Advisor. Product Data Sheet October 2014
Product Data Sheet Energy Advisor Reduce energy consumption and increase energy visibility with Energy Advisor Manage energy targets with up-to-the minute information across multiple Energy Account Centers
More informationOP3 INTEGRATES ORACLE E-BUSINESS SUITE PROJECTS AND
OP3 INTEGRATES ORACLE E-BUSINESS SUITE PROJECTS AND KEY BENEFITS Seamless business flows across the lifecycle of your projects. Seamless best-of-breed system ensures the same information is obtained and
More informationThe retailer is the final frontier of
Demand planning and Forecasting with Pos Data: A Case Study By Fred Andres The retailer is the final frontier of supply chain planning. So, it is important for manufacturers to have a serious look at what
More informationCollecting Clothing Data from the Internet
Collecting Clothing Data from the Internet Robert Griffioen, Jan de Haan and Leon Willenborg Statistics Netherlands Aim of the paper Our aim is to describe Statistics Netherlands experiences with the collection
More informationThe Big Business of Big Data in Liquidity Management How Big Data Can Revolutionize Your Cash Forecast
The Big Business of Big Data in Liquidity Management How Big Data Can Revolutionize Your Cash Forecast Mid Atlantic Association for Financial Professionals (MAAFP) Mark O Toole @MarkDOToole Big data? What
More informationPeak ETA Product Overview
Oracle Business Intelligence 11g Peak ETA Product Overview Antony Heljula / Paul Oprea January 2013 Peak Indicators Limited 2 Peak ETA Product Overview Agenda What is Peak ETA? Real Business Value 2 Examples
More informationThe Riksbank s Business Survey. Industrial activity stronger than expected
The Riksbank s Business Survey Industrial activity stronger than expected February 2017 THE RIKSBANK S BUSINESS SURVEY FEBRUARY 2017 3 The Riksbank s Business Survey in February 2017 1 According to Swedish
More informationPrice dispersion in the Internal Market 1
Price dispersion in the Internal Market 1 How far are we from reaching the limits of price convergence in the Internal Market? Introduction Price comparisons are good indicators of integration and market
More informationGADD Analytics Overview
GADD Analytics Overview Version 5.4 A GADD White Paper Published May, 2012 www.gaddsoftware.com Table of content 1. Introduction... 3 2. What is GADD Analytics?... 3 2.1. What are the benefits with GADD
More informationCan Advanced Analytics Improve Manufacturing Quality?
Can Advanced Analytics Improve Manufacturing Quality? Erica Pettigrew BA Practice Director (513) 662-6888 Ext. 210 Erica.Pettigrew@vertexcs.com Jeffrey Anderson Sr. Solution Strategist (513) 662-6888 Ext.
More informationSupply Chain MICROSOFT BUSINESS SOLUTIONS DEMAND PLANNER
Supply Chain MICROSOFT BUSINESS SOLUTIONS DEMAND PLANNER DEMAND PLANNING FOR BUSINESSES Demand planning is the first step towards business planning. As businesses are moving towards a demand-centric environment
More informationWatch our predictions come true! In-Database Analytics with EXASOL: Unlocking Big Data Potential in E-Commerce
Watch our predictions come true! Case Study // In-Database Analytics with EXASOL: Unlocking Big Data Potential in E-Commerce Management Summary Keywords Online shopping Market basket analysis Data analysis
More informationWHITE PAPER. Integrated customer insights
WHITE PAPER Integrated customer insights Customer centricity is the new voice for brands the world over. Every brand will have online and offline customers, depending on the touch points they wish to deal
More informationUntil the start of 2007, the volume series of the quarterly
From constant prices to chain-linked prices: what consequences? Franck Arnaud Guillaume Mordant Bertrand Lhommeau Christelle Minodier Franck Cachia Frédéric Tallet Divisions - Quarterly National Accounts
More informationASSOCIATION EUROPEENNE DU COMMERCE DE FRUIT ET LEGUMES, AISBL EUROPEAN FRUIT AND VEGETABLE TRADE ASSOCIATION, AISBL
ASSOCIATION EUROPEENNE DU COMMERCE DE FRUIT ET LEGUMES, AISBL EUROPEAN FRUIT AND VEGETABLE TRADE ASSOCIATION, AISBL Brussels February 28, 2008 EUCOFEL S CONTRIBUTION TO THE EC CONSULTATION ON SCHOOL FRUIT
More informationENTERPRISE SIMULATION A PRACTICAL APPLICATION IN BUSINESS PLANNING. Robert Suggs Brian Lewis
Proceedings of the 2007 Winter Simulation Conference S. G. Henderson, B. Biller, M.-H. Hsieh, J. Shortle, J. D. Tew, and R. R. Barton, eds. ENTERPRISE SIMULATION A PRACTICAL APPLICATION IN BUSINESS PLANNING
More informationPOWER BI BOOTCAMP. COURSE INCLUDES: 4-days of instructor led discussion, Hands-on Office labs and ebook.
Course Code: AUDIENCE: FORMAT: LENGTH: POWER BI BOOTCAMP O365-410-PBI (CP PBI365) Data Analyst and BI Professionals Instructor-led training with hands-on labs 4 Days COURSE INCLUDES: 4-days of instructor
More informationIBM Planning Analytics Express
Performance management and business intelligence for midsize organisations IBM Planning is a performance management (PM) and business intelligence (BI) solution for midsize organisations. It delivers the
More informationIBM Kenexa Talent Insights
IBM Kenexa Talent Insights A rich source of analytics capabilities from IBM s investment 2 Driving business performance Cognitive Integrated Predictive Prescriptive Descriptive 3 How do you get insights
More informationSECANT Desktop can be configured to produce patterns with or without turns, either including or prohibiting third phase cutting / recuts.
1.0 Overview Whether you are cutting sheet metal, wood, board, plastic, or glass, SECANT Desktop provides fast and accurate production of low wastage cutting patterns for all beam saw operations. SECANT
More informationConfigurable Policy Enforcement. Automated Remedy Actions. Granular Reporting - Scheduled and On-Demand
LayerX Technologies, is a leading provider of advanced data analytics software for the IT industry. Our solutions are used across multiple IT domains to provide rich insight into application performance
More informationAn Assessment of the ISM Manufacturing Price Index for Inflation Forecasting
ECONOMIC COMMENTARY Number 2018-05 May 24, 2018 An Assessment of the ISM Manufacturing Price Index for Inflation Forecasting Mark Bognanni and Tristan Young* The Institute for Supply Management produces
More informationResearch indices using web scraped price data: clustering large datasets into price indices (CLIP)
Article Research indices using web scraped price data: clustering large datasets into price indices (CLIP) Web scraping is a tool for extracting information from the underlying HTML code of websites. ONS
More informationConcept Searching is unique. No other statistical search and classification vendor puts compound terms in their index. This technique delivers high
1 2 Concept Searching is unique. No other statistical search and classification vendor puts compound terms in their index. This technique delivers high precision without the loss of recall. Why Classification?
More informationThe Power of Digital Printing
September 2002 The Power of Digital Printing Taking the next step to new opportunity and profitability Jim Hamilton Associate Director CAP Ventures, Inc. Page 3 Contents 3 Straight Talk About Digital Print
More information3 CRITICAL STEPS TO MAKE YOUR RPA IMPLEMENTATION A SUCCESS
WHITEPAPER 3 CRITICAL STEPS TO MAKE YOUR RPA IMPLEMENTATION A SUCCESS www.celonis.com IN THIS WHITEPAPER Robotic Process Automation (RPA) is quickly becoming a must-have initiative for modern businesses.
More informationASSIGNMENT SUBMISSION FORM
Course Name: Assignment Title: Submitted by: ASSIGNMENT SUBMISSION FORM Treat this as the first page of your assignment FCAS Group Member Name Sanchit Garg 61310634 Ankur Pandey 61310573 Mohammad Shahid
More informationBUILDING OMNI-CHANNEL RETAIL FROM THE BACK END UP
BUILDING OMNI-CHANNEL RETAIL FROM THE BACK END UP Building Omni-Channel Retail From The Back End Up 2 An average of 50% of retailers plan to increase their commitments to order management, e-commerce,
More informationThe State of Cross-Device Commerce
The State of Cross-Device Commerce 1 STATE OF CROSS-DEVICE COMMERCE H2 2016 Copyright 2017 Criteo H2 2016 United Kingdom The challenge ahead Closing the gap across devices A new paradigm. Marketing personalisation
More informationBaltic Marine Environment Protection Commission
Baltic Marine Environment Protection Commission Expert Working Group for Mutual Exchange and Deliveries of AIS & Data Stockholm, Sweden, 7-8 June 2017 AIS EWG 28-2017 Document title Draft methodology to
More informationContents Getting Started... 9 Sample Dashboards... 17
Analytics Reference Guide 16 R1 March 2016 Contents Getting Started... 9 About Oracle Primavera Analytics... 10 Prerequisites to Use Primavera Analytics... 11 About Analyses... 11 About s... 12 About
More informationThe implications of Big Data for BTS and COS George Kershoff
The implications of Big Data for BTS and COS George Kershoff Presented at the 7th joint EC-OECD workshop on Recent developments in Business and Consumer Surveys held in Paris on 30 November and 1 December
More informationMulti-Touch Attribution
Multi-Touch Attribution BY DIRK BEYER HEAD OF SCIENCE, MARKETING ANALYTICS NEUSTAR A Guide to Methods, Math and Meaning Introduction Marketers today use multiple marketing channels that generate impression-level
More informationORACLE HYPERION PLANNING
ORACLE HYPERION PLANNING KEY FEATURES AND BENEFITS KEY FEATURES: Multi-dimensional / multi user planning with a powerful business rules engine Flexible workflow and plan management capabilities Comprehensive
More informationEIS and Data Warehousing applications in the Telecommunications Industry
EIS and Data Warehousing applications in the Telecommunications Industry Garth Brown, Mobile Telephone Networks (MTN), Johannesburg, South Africa Background to MTN EIS and Data Warehouse Project. The MIS
More informationSelecting the Right SAP BusinessObjects BI Client Product Based on Your Business Requirements for SAP BW Customers
Product Based on Your Business Requirements for SAP BW Customers Ingo Hilgefort Director Solution Management Disclaimer This presentation outlines our general product direction and should not be relied
More informationAdding Apples and Oranges: Keeping Score with Gross Domestic Product. At this very moment at our house, we have three apples, two oranges, four
Adding Apples and Oranges: Keeping Score with Gross Domestic Product At this very moment at our house, we have three apples, two oranges, four bananas and an overripe pear in the fruit basket on top of
More informationChapter 8 Analytical Procedures
Slide 8.1 Principles of Auditing: An Introduction to International Standards on Auditing Chapter 8 Analytical Procedures Rick Hayes, Hans Gortemaker and Philip Wallage Slide 8.2 Analytical procedures Analytical
More informationMonitoring consumer price trend using daily price data of online grocery stores in India
Asia-Pacific Economic Statistics Week Seminar Component Bangkok, 2 4 May 2016 Name of author Abhiman Das (Professor, IIM Ahmedabad) Pulak Ghosh (Professor, IIM Bangalore) Anirban Sanyal (Research Officer,
More informationMachine Learning Technologies for The Hospitality Industry
Machine Learning Technologies for The Hospitality Industry The emergence of new computing technologies has unlocked the gate for hospitality businesses to enter the Machine Learning and Big Data arena.
More information36.2. Exploring Data. Introduction. Prerequisites. Learning Outcomes
Exploring Data 6. Introduction Techniques for exploring data to enable valid conclusions to be drawn are described in this Section. The diagrammatic methods of stem-and-leaf and box-and-whisker are given
More informationRecruitment and selection toolkit. Guide to recruitment advertising
Recruitment and selection toolkit Guide to recruitment advertising Contents 2 1 Introduction Purpose of the guide 3 2 Drafting the advert Roles & responsibilities 4 Advert content 4-6 ERNITY & ADOPTI 3
More informationReview: Simple schedule risk modelling with Safran Risk
Creating value from uncertainty Broadleaf Capital International Pty Ltd ABN 24 054 021 117 www.broadleaf.com.au Review: Simple schedule risk modelling with Safran Risk With a view to exploring alternative
More informationTerms of Reference: Short-term Advisory Mission for World Service Data Management and IT architecture
Terms of Reference: Short-term Advisory Mission for World Service Data Management and IT architecture Background As LWF World Service s project and funding portfolio continues to grow and diversify, there
More informationThe State of Cross-Device Commerce
The State of Cross-Device Commerce 1 STATE OF CROSS-DEVICE COMMERCE H2 2016 Copyright 2017 Criteo H2 2016 Australia The challenge ahead Closing the gap across devices A new paradigm. Marketing personalisation
More informationImproving enterprise performance through operations intelligence solutions siemens.com/xhq
XHQ Operations Intelligence Software Improving enterprise performance through operations intelligence solutions siemens.com/xhq What is XHQ? That s simple. XHQ is a better, more intelligent way to make
More informationEuropean Commission. Replacing judgement by statistics: New Consumer Confidence Indicators
European Commission Directorate General Economic and Financial Affairs Replacing judgement by statistics: New Consumer Confidence Indicators Andreas Reuter Business and consumer surveys and short-term
More informationUncover possibilities with predictive analytics
IBM Analytics Feature Guide IBM SPSS Modeler Uncover possibilities with predictive analytics Unlock the value of data you re already collecting by extracting information that opens a window into customer
More informationBest Practices in Demand and Inventory Planning
W H I T E P A P E R Best Practices in Demand and Inventory Planning for Chemical Companies Executive Summary In support of its present and future customers, CDC Software sponsored this white paper to help
More informationTransparent Quality Reporting in the Integration of Multiple Data Sources: Quality of Input Data
Transparent Quality Reporting in the Integration of Multiple Data Sources: Quality of Input Data Mark Prell Economic Research Service FCSM Research Conference March 7, 2018 The views expressed are those
More informationQlik Sense. Data Sheet. Transform Your Organization with Analytics
Data Sheet Qlik Sense Transform Your Organization with Analytics Qlik Sense is Qlik s next-generation platform for modern, self-service oriented analytics. It supports the full range of analytics use cases
More informationHow Much Do We Know About Savings Attributable to a Program?
ABSTRACT How Much Do We Know About Savings Attributable to a Program? Stefanie Wayland, Opinion Dynamics, Oakland, CA Olivia Patterson, Opinion Dynamics, Oakland, CA Dr. Katherine Randazzo, Opinion Dynamics,
More informationWhy Fishbowl Manufacturing and Fishbowl Warehouse Are #1 Among QuickBooks Users
Why Fishbowl Manufacturing and Fishbowl Warehouse Are #1 Among QuickBooks Users Fishbowl is the most popular manufacturing and warehouse management software for QuickBooks users. Fishbowl Manufacturing
More information1.0 Chapter Introduction
1.0 Chapter Introduction In this chapter, you will learn to use price index numbers to make the price adjustments necessary to analyze price and cost information collected over time. Price Index Numbers.
More informationA STATISTICAL ANALYSIS OF REVISIONS OF SWEDISH NATIONAL ACCOUNTS DATA *
Finnish Economic Papers Volume 28 Number 1 Fall 217 A STATISTICAL ANALYSIS OF REVISIONS OF SWEDISH NATIONAL ACCOUNTS DATA * CAROLINE FLODBERG ** Sveriges Riksbank and PÄR ÖSTERHOLM *** School of Business,
More informationCooladata Delivers a Competitive Edge for the Webcollage Platform
The Story Cooladata Delivers a Competitive Edge for the Webcollage Platform 2 Webcollage is the leading cloud-based, rich product content syndication platform offering powerful analytics and behavioral
More informationOUTLINE. 1. Focus of the CIPR. 2. Web Presence. 3. The Archival Process. 4. Public Data Archives. 5. Work-in-progress
OUTLINE 1. Focus of the CIPR 2. Web Presence 3. The Archival Process 4. Public Data Archives 5. Work-in-progress OUR FOCUS Public data and research dissemination portal. Compiles absolute (domestic currency)
More informationForecasting fruit demand: Intelligent Procurement
2012 Forecasting fruit demand: Intelligent Procurement FCAS Final Project Report Predict fruit sales for a 2 day horizon to efficiently manage procurement logistics. Dinesh Ganti(61310071) Rachna Lalwani(61310845),
More informationElevate your brand online
Elevate your brand online The amazing opportunity for retailers to put their brand in front of potential customers ready to buy. Contents This E-Book will explore how the consumer buying journey has changed
More informationNew era of data analysing. Product Presentation 2014
New era of data analysing Product Presentation 2014 Trace Analyzer overview Trace Analyzer is a tool for post analyzing of massive data amounts from various different sources. It reduces the time used
More informationCHAIR Welcome. Would you like to make an opening statement?
E 102 Senate Tuesday, 6 October 2009 CHOU, Associate Professor Chun Tung, Private capacity KANHERE, Dr Salil, Private capacity Evidence was taken via teleconference CHAIR Welcome. Would you like to make
More informationChart your future with predictive analytics
IBM Analytics Feature Guide IBM SPSS Modeler Chart your future with predictive analytics Finding hidden trends in your data can give you tremendous insights into your business. 2 Chart Your Future Contents
More informationFood and Beverage Companies Become Market Leaders with Vanguard Predictive Planning
Food and Beverage Companies Become Market Leaders with Vanguard Predictive Planning The food and beverage industry is not for the faint of heart. A company s success depends heavily on its ability to effectively
More informationIs More Always Better? A Comparison of Billing Regression Results Using Monthly, Daily and Hourly AMI Data
Is More Always Better? A Comparison of Billing Regression Results Using Monthly, Daily and Hourly AMI Data John Cornwell, Evergreen Economics, Portland, OR Stephen Grover, Evergreen Economics, Portland,
More informationEfficient Troubleshooting Using Machine Learning in Oracle Log Analytics
Efficient Troubleshooting Using Machine Learning in Oracle Log Analytics Nima Haddadkaveh Director, Product Management Oracle Management Cloud October, 2018 Safe Harbor Statement The following is intended
More informationSAS Decision Manager
SAS Decision Manager A Technical Supplement James Taylor CEO SAS Decision Manager combines business rules management with predictive analytic models and analytic model management. These capabilities are
More informationReprinted from. A supplement to Hydrocarbon Engineering. Spring
Reprinted from A supplement to Hydrocarbon Engineering Spring 2010 www.cbi.com Martin te Lintelo, Yokogawa Europe B.V., The Netherlands, discusses high level control system design for LNG receiving terminals.
More informationImplementation of the treatment of the scanner data in France
Implementation of the treatment of the scanner data in France Guillaume Rateau Head of the CPI methodology section Introduction : project schedule since 2009 Solution studies, implementation of the IT
More informationBPIC 13: Mining an Incident Management Process
BPIC 13: Mining an Incident Management Process Emmy Dudok, Peter van den Brand Perceptive Software, 22701 West 68th Terrace, Shawnee, KS 66226, United States {Emmy.Dudok, Peter.Brand}@perceptivesoftware.com
More informationon RETAIL MERCHANDISING EFFICIENCY
October 1962 A.E. 338 EFFECT of PACKAGING PRODUCE on RETAIL MERCHANDISING EFFICIENCY Department of Agricultural Economics and Rural Sociology Ohio Agricultural Experiment Station Wooster, Ohio EFFECTS
More informationAmazing Experiences With Launchpads In SharePlus
Amazing Experiences With Launchpads In SharePlus Amazing Experiences With LaunchPads In SharePlus Trend: The Mobile User Experience The trend is here to stay. Mobile devices are everywhere, giving executives
More informationForecasting Tax Revenue at the U.S. Department of the Treasury with FocalPoint TaxInsight. Leading the way in tax analytics
Forecasting Tax Revenue at the U.S. Department of the Treasury with FocalPoint TaxInsight Leading the way in tax analytics Accurate forecasting of tax revenues can help Federal and State Government Agencies
More informationBest practices in demand and inventory planning
whitepaper Best practices in demand and inventory planning WHITEPAPER Best Practices in Demand and Inventory Planning 2 about In support of its present and future customers, Aptean sponsored this white
More informationProduct comparison: Prophix Enterprise vs. PROPHIX 10 R3
Product comparison: vs. Prepared by Software Purpose This document is intended for current customers who are considering migrating to. Introduction is an integrated application that is built on an open
More informationPowered by. IoT STRATEGY : INSIGHTS FROM EARLY IoT ADOPTERS
Powered by IoT STRATEGY : INSIGHTS FROM EARLY IoT ADOPTERS 2 GLOBAL IoT REPORT 2017 Executive Summary Technology and trends come and go, but IoT has what it takes to transform organizations and shape the
More informationDatameer for Data Preparation: Empowering Your Business Analysts
Datameer for Data Preparation: Empowering Your Business Analysts As businesses strive to be data-driven organizations, self-service data preparation becomes a critical cog in the analytic process. Self-service
More informationCisco s Digital Transformation Supply Chain for the Digital Age
Cisco s Digital Transformation Supply Chain for the Digital Age The Cisco Supply Chain: Global, Complex, and Diverse Cisco s global supply chain extends across 13 countries and more than 25 locations.
More informationProactive Listening: Taking Action to Meet Customer Expectations
2013 Proactive Listening: Taking Action to Meet Customer Expectations Summary: Proactive listening lets your company meet your customers demands for coordinated, responsive treatments. A proactive listening
More informationWhite Paper. Transforming Contact Centres using Simulation-based Scenario Modelling
White Paper Transforming Contact Centres using Simulation-based Scenario Modelling Meet Your KPI s, Deliver Consistently High Service and Reduce Customer Churn to Increase Your Bottom Line Results PM@SIMUL8.com
More informationRedesigning Economic Statistics at Statistics Finland
1(9) Redesigning Economic Statistics at Statistics Finland Introduction Statistical description is challenged by the changes in the economy and the growing demand for more and better statistics with less
More information